复杂人机共融场景中人体姿态识别及避碰策略综述

智能机器人与人类智慧的融合，即人机协作共融，已经实现了将机器人的机械优势和人类的高级认知能力集中于同一个工作架构之中，能够在复杂环境中协同作业，从而提高效率。针对复杂的人机共融场景，特别是机器人在诸如光线条件变化、背景干扰以及运动过程，对比总结了基于机器视觉的人体姿态识别方法和基于机器学习的避碰策略，详细比较各类方法的研究现状及应用，并探讨了基于深度学习的目标识别和避碰方法的发展及应用。

A Review of Human Posture Recognition and Collision Avoidance Strategies in Complex Human-robot Collaboration Scenarios

The integration of intelligent robots with human intelligence, that is, the harmonious collaboration between humans and robots, has been realized. The mechanical advantages of robots and the advanced cognitive abilities of humans are combined within the same working framework. This collaboration enables them to work together in complex environments to improve efficiency. For complex human-robot integration scenarios, especially when robots encounter varying lighting conditions, background interference, and during movement, a comparison has been made between human posture recognition methods based on machine vision and collision avoidance strategies based on machine learning. The current status and applications of various methods have been detailedly compared, and the development and application of object recognition and collision avoidance methods based on deep learning have been discussed.